[Published by OECD Publishing] Power Generation (BookFi.org)

POWER GENERATIONINVESTMENT IN

ELECTRICITY MARKETS

Most IEA countries are liberalising their electricity markets,shifting the responsibility for financing new investment in powergeneration to private investors. No longer able to automaticallypass on costs to consumers, and with future prices of electricity

uncertain, investors face a much riskier environment forinvestment in electricity infrastructure.

This report looks at how investors have responded to the needto internalise investment risk in power generation. While capital

and total costs remain the parameters shaping investment choices,the value of technologies which can be installed quickly and

operated flexibly is increasingly appreciated. Investors are alsomanaging risk by greater use of contracting, by acquiring retail

businesses, and through mergers with natural gas suppliers.

While liberalisation was supposed to limit government interventionin the electricity market, volatile electricity prices have put pressureon governments to intervene and limit such prices. This study looks

at several cases of volatile prices in IEA countries electricity markets,and finds that while market prices can be a sufficient incentive for

new investment in peak capacity, government intervention intothe market to limit prices may undermine such investment.

P

o

w

e

r

G

e

n

e

r

a

t

i

o

n

I

n

v

e

s

t

m

e

n

t

i

n

E

l

e

c

t

r

i

c

i

t

y

M

a

r

k

e

t

s

-:HSTCQE=VUZZ[U:(61 2003 30 1 P1) 92-64-10556-5 75

2003

I N T E R N AT I O N A LE N E RGY AG E N CY

ENERGYMARKETREFORM

POWERGENERATIONINVESTMENT

IN ELECTRICITYMARKETS

Powergene3 7/10/03 16:46 Page 1

I N T E R N AT I O N A LE N E RGY AG E N CY

ENERGYMARKETREFORM

POWERGENERATIONINVESTMENT

IN ELECTRICITYMARKETS

OECD/IEA, 2003

Applications for permission to reproduce or translate all or part of this publication should be made to:Head of Publications Service, OECD/IEA

2, rue Andr-Pascal, 75775 Paris Cedex 16, Franceor

9, rue de la Fdration, 75739 Paris Cedex 15, France.

INTERNATIONAL ENERGY AGENCY9, rue de la Fdration,

75739 Paris Cedex 15, France

The International Energy Agency (IEA) is anautonomous body which was established inNovember 1974 within the framework of theOrganisation for Economic Co-operation andDevelopment (OECD) to implement an inter-national energy programme.

It carries out a comprehensive programme ofenergy co-operation among twenty-six* of the

of the IEA are:

to maintain and improve systems for copingwith oil supply disruptions;

to promote rational energy policies in a globalcontext through co-operative relations with non-member countries, industry and internationalorganisations;

to operate a permanent information system onthe international oil market;

to improve the worlds energy supply anddemand structure by developing alternativeenergy sources and increasing the efficiency ofenergy use;

to assist in the integration of environmental andenergy policies.

*Belgium, Canada, the Czech Republic, Denmark,Finland, France, Germany, Greece, Hungary, Ireland,Italy, Japan, the Republic of Korea, Luxembourg,the Netherlands, New Zealand, Norway, Portugal,

UnitedKingdom, the United States. The EuropeanCommission also takes part in the work of the IEA.

ORGANISATION FOR ECONOMIC CO-OPERATION

AND DEVELOPMENT

Pursuant to Article 1 of the Convention signed inParis on 14th December 1960, and which cameinto force on 30th September 1961, the Organisationfor Economic Co-operation and Development(OECD) shall promote policies designed:

to achieve the highest sustainable economicgrowth and employment and a rising standard

financial stability, and thus to contribute to thedevelopment of the world economy;

to contribute to sound economic expansion in

process of economic development; and

to contribute to the expansion of world tradeon a multilateral, non-discriminatory basis inaccordance with international obligations.

Germany, Greece, Iceland, Ireland, Italy,Luxembourg, the Netherlands, Norway, Portugal,Spain, Sweden, Switzerland, Turkey, the UnitedKingdom and the United States. The following

through accession at the dates indicatedhereafter: Japan (28th April 1964), Finland(28th January 1969), Australia (7th June 1971),New Zealand (29th May 1973), Mexico (18thMay 1994), the Czech Republic (21st December1995), Hungary (7th May 1996), Poland (22ndNovember 1996), the Republic of Korea (12thDecember 1996) and Slovakia (28th September2000). The Commission of the EuropeanCommunities takes part in the work of the OECD(Article 13 of the OECD Convention).

OECDs thirty member countries. The basic aims

member as well as non-member countries in the

of living in member countries, while maintaining

Austria, Belgium, Canada, Denmark, France,The original member countries of the OECD are

countries became members subsequently

Spain, Sweden, Switzerland, Turkey, the

IEA member countries: Australia, Austria,

FOREWORDElectricity policy in OECD countries over the past decade has been focused onthe liberalisation of electricity markets. In doing so, governments have shiftedthe responsibility for financing investment in power generation away fromgenerally state-owned monopolies to private investors. No longer able toautomatically pass on costs to consumers, and with future prices of electricityuncertain, investors face a much riskier environment for power generationinvestment decisions.

Governments have remained concerned about the adequacy and thecomposition of power generation investment in a liberalised market. PreviousIEA work has found that in most markets which were opened to competition,the level of investment in generating capacity has been adequate.This reportlooks at how internalising the risks of investment has changed the investors andthe composition of investments. The report finds that while capital and totalcosts remain the fundamental parameters shaping investment choices,technologies which can be installed quickly and operated flexibly in response tomarkets are increasingly attractive. Power firms are also responding to theuncertain environment by finding ways to mitigate their investment risks, bygreater use of contracting, by acquiring retail businesses, and through mergerswith natural gas supply businesses.

Relying on market prices to signal the need for new power generationinvestment and to stimulate a timely reaction of the investors, particularly forpeaking capacity, has introduced a new challenge for governments. Whileliberalisation was expected to limit government intervention in the electricitymarket, volatile electricity prices, signalling the need for greater capacity, haveraised prices to consumers and put pressure on governments to intervene.Thisstudy looks at several cases of volatile prices in IEA electricity markets, thegovernments responses to these prices and the implications for investment inpower generation, particularly in peak capacity. It finds that market prices canbe a sufficient incentive for new investment in peak capacity, and thatgovernment intervention into the market only to limit prices may underminesuch investment.

Yet there remains much for governments to do. It is vital that governmentsdefine the roles of all players in reformed electricity markets as clearly as

3

POWER GENERATION INVESTMENT IN ELECTRICITY MARKETS

POWER GENERATION INVESTMENT IN ELECTRICITY MARKETS x

possible. For governments, that means establishing the policies and leaving theirimplementation to regulatory authorities, and the pricing and the reaction toprices to the market players. In particular, it is for governments to define theirobjectives for the electricity sector, and for the regulator to design the rules, sothat sufficient production capacity or responsive demand can be properlyfinanced, with an adequate return on investment. In addition, governmentsshould ensure that, when a decision to invest is taken, its implementation is notdelayed by excessive or inappropriate authorisation procedures. Moreimportantly, to attract investment, governments must maintain theircommitment to ensuring the success of the market, and not reverse directionunder the pressure of price hikes or other unpleasant surprises along the way.Certainty over the direction that legislation will take is one of the mostimportant preconditions for investors.

This book is part of the IEA series on electricity market reform. It is publishedunder my authority as Executive Director of the International Energy Agency.

Claude MandilExecutive Director

4

ACKNOWLEDGEMENTSThe principal author of this report is Peter Fraser of the EnergyDiversification Division, working under the direction of Ralf Dickel, Headof Division, and No van Hulst, Director of the Office for Long Term Co-operation and Policy Analysis.

This book has benefited greatly from the papers presented at an IEA/NEAworkshop, Power Generation Investment in Liberalised ElectricityMarkets held at the IEA on 25-26 March 2003.The comments receivedfrom the IEAs Standing Committee on Long Term Co-operation, DougCooke and Maria Argiri from the IEA Secretariat, John Paffenbarger ofConstellation Energy, and Dean Travers of Electrabel have all been veryhelpful.

5


TABLE OF CONTENTS

EXECUTIVE SUMMARY 11

INTRODUCTION 21

RISK AND POWER GENERATION INVESTMENT 27

HEDGING RISK IN POWER GENERATION INVESTMENT 43

PRICE VOLATILITY, INVESTMENT AND GOVERNMENT INTERVENTION 55

CONCLUSIONS AND RECOMMENDATIONS 89

BIBLIOGRAPHY 95

LIST OF TABLES

1. Summary of Markets Experiencing Electricity Price Crises 26

2. Qualitative Comparison of Generating Technology by Risk Characteristics 32

3. Comparison of Levelised Unit Energy Cost Estimates Using DifferentMethodologies 36

4. Comparison of European Electricity Market Liquidity (as of April 2003) 45

5. World Top 15 Deals in Gas-Electricity Cross-Sectoral Mergers andAcquisitions 50

5

4

3

2

1

7


6. Prices Needed to Recoup Long-Run Marginal Cost in PJM in Top 5% of Hours by Load 57

7. Markets Experiencing Electricity Price Crises 60

8. Electricity Generation in the Australian Market (2000/2001) 65

9. Major Power Stations Placed in Service in South Australia 68

10. Costs for Model New Entry Generating Technologies Serving South Australia 68

11. Net Earnings of Model New Entrant Power Plants in South Australia 70

12. Impact of AUD 1500/MWh Price Cap on Peaking Plant Profitability 71

13. Changes in Retail Electricity Prices in Norway (2003 Q1 vs. 2002 Q1) 75

LIST OF FIGURES

1. Global Power Generation Orders, Including Gas Turbines (1950-2003) 23

2. Share of Projected Capacity Additions in OECD Generating Capacity by Fuel (1999-2030) 24

3. Net Present Value Frequency Range for CCGT Investment 37

4. Growth of Electricity Trade in Nordic Electricity Markets (1998-2003) Elspot Market 46

5. Growth of Electricity Trade in Nordic Electricity Markets (1998-2003) Financial Markets 47

6. Generating Capacity of the Largest EU-15 Utilities (2002) 52

7. Spot Electricity Prices in the Victoria Market (July 1999 June 2003) 56

8. Wholesale Prices in the Ontario Electricity Market (May 2002 April 2003) 62

9. Retail Residential and Wholesale Prices in Alberta, Canada (1998-2002) 64

10. States in the Australian National Electricity Market 65

11. Monthly Average Spot Prices in Victoria/South Australia Electricity Markets (December 1998 to May 2003) 66

8

POWER GENERATION INVESTMENT IN ELECTRICITY MARKETS x

12. Distribution of South Australia Spot Power Prices (1999-2002) 69

13. Monthly Averages of NZ Spot Electricity Prices (August 1999 July 2003) 72

14. Nord Pool Weekly Averages of Spot Electricity Prices (2003 vs. 2002) 75

15. Comparison of Norway Household Electricity Prices(variable prices vs. one-year contract) 76

16. Demand Response in Electricity Markets 81

17. Scatter Plot of Prices vs. Demand in the Ontario Market (May 2002 April 2003) 82

18. Demand Curve for Capacity Auction for New York ISO (New York City) 88

9


EXECUTIVE SUMMARY

Electricity markets have been opened to competition in nearly all IEAcountries. The introduction of competition in the generation and retailsupply of electricity should improve the economic efficiency of the powersector. There is a great deal of evidence to suggest that marketliberalisation has generally improved the productive efficiency of thesector. Furthermore, with many markets opened in the presence ofsurpluses of generating capacity, competitive pressures have reducedpower prices.

The key long-term question for improving the economic efficiency of theelectricity sector is whether investment performance in powergeneration will improve. Much of the focus has been on whether theoverall level of investment in generating capacity has been adequate toensure security of electricity supply. Previous IEA work1 has shownsubstantial investment in power generation has taken place and thatOECD electricity markets generally provide reliable electricity supplywith the exception of California notwithstanding2. However, it alsoconcluded that the biggest challenges remain ahead most markets arejust beginning to approach their first major investment cycle as surpluscapacity is absorbed.

This report examines three other major issues associated with powergeneration investment in liberalised electricity markets.The first issue ishow liberalisation and investment risk have affected technology choicesin power generation. In particular, given the anticipated growth ininvestment in natural gas-fired power generation, it asks whether short-termism by investors is undermining an economically rational allocationof capital. Chapter 2 examines the risks that investors in powergeneration face in liberalised markets.

The second issue addressed is how power generation investors areadapting to investment risks. Financial hedges have not developed

11

EXECUTIVE SUMMARY

1. Security of Supply in Electricity Markets: Evidence and Policy Issues, International Energy Agency, Paris 2002.

2. This report also concluded that existing transmission interconnection capacity was insufficient, and there was a need forpolicies to encourage such investment.

sufficiently to offer significant long-term hedging value to investors.As aconsequence, firms are seeking to hedge these risks by the use of long-term contracts, by integration of generation with retail, by growing in size,and by mergers between gas and electricity companies. Chapter 3 looksat these strategies and discusses whether such structural changes areconsistent with efficient energy markets.

The emergence of tighter electricity markets has led to increasedelectricity price volatility. Increased price volatility has put pressure ongovernments to intervene in electricity markets. The report reviewsrecent cases of electricity price volatility in Norway, Canada, Australiaand New Zealand and discusses the impact of the governmentsresponse to these crises on the prospects for efficient new generationinvestment.

Risk and its Influence on Power GenerationInvestment and Corporate Structure

The reform of electricity and gas markets has led to major changes in theway decisions are taken on power sector investment. Opening the sectorto competition has led to the internalisation of risk in investmentdecision-making. Investors now examine power generation optionsaccording to the different financial risks posed by the differenttechnologies.

Given the long-term nature of electricity investments, investmentdecisions in baseload generating capacity are being made on the basis oflong-term fundamentals rather than looking at short-term behaviour inthe spot or forward electricity markets. Conventional discounted cashflow methods are still most often used. Nevertheless, investors arebeginning to take account of differences in risk levels in assessing thelikely profitability of different investments.

The current market preference for gas-fired power generation forbaseload generation in many OECD countries can be explained mainly bythe perceived lower cost of gas-fired generation. The characteristics ofthe combined cycle gas turbine (CCGT), its low capital cost, and itsflexibility have also added to its attractiveness.The importance of CCGT

12

EXECUTIVE SUMMARY x

means that gas markets assume a greater importance than ever forpower generation development. For governments, this means movingforward on liberalisation, and monitoring investment in both gas andelectric infrastructure.

The preference for gas-fired power generation does expose investors toincreased fuel price risk.The creation or development of electricity andnatural gas markets has led to a system where, in the absence of hedgingpossibilities, price risks cannot be managed, but must be assessed byprobabilistic approaches.

Nuclear plants can be financed in an electricity market, but this will bemuch easier if customers share the risks.The decision of Finlands TVOcompany, a large electricity co-operative (see Chapter 3), to proceed withthe investment in a nuclear plant is the first by a company in a competitiveelectricity market. The structure of the investment is exceptional,however, with large consumers of electricity willing to take the risks ofinvesting in a nuclear plant because they expect to be able to obtain thelong-term financial benefits.

The adequacy of investment in peaking generation remains an even moresensitive issue.The low numbers of hours of operation of peaking planthave led some to question whether markets can bring forward adequatepeaking capacity. Low capital and high flexibility in operation areparticularly important attributes in attempting to value peaking capacity.While investors have long been aware of the qualitative benefits offlexibility in a liberalised market, new techniques have now beendeveloped to quantify the value of flexibility. In particular, they suggestthat flexible peaking capacity may be much more profitable thantraditional approaches would predict.

The current economic downturn in the US electricity market has lenderslooking for strong companies with stable revenue flows and customerbases for future investments. Liberalisation has also affected the waypower plants are financed. Early enthusiasm for the merchant power plantmodel, where power plants are financed without the security of regulatedprofits, has dissipated thanks to adverse investment experience in theUnited States.

13

EXECUTIVE SUMMARY

Hedging Risk in Power Generation Investment

Firms are also seeking new strategies to hedge their investments. Theabsence of liquidity in financial markets for electricity and for natural gas,particularly for longer-term products, means that investors must seekother means of hedging their electricity price (and fuel price) risks.Longer-term contracts for electricity between producers and retailershelp both manage price volatility.There has also been increased mergeractivity between generators and retailers in several markets. Mergersbetween electricity and gas producers are partially motivated byincreased opportunities for arbitrage.

Power generating companies are also growing in size, in part to be ableto finance more of their investment on their own balance sheets.Thesemergers are increasing concerns about the potential impacts oncompetition in electricity markets and can be expected to continue toattract close regulatory attention.

Electricity Markets, Electricity Price Volatility and Investment

Chapter 4 examines recent price crises in Australia (South Australia),Canada (Ontario and Alberta), Norway and New Zealand. Drawinggeneral lessons from government management of electricity pricecrises and its impact on investment is difficult given the rather differentcircumstances of each crisis and the relatively short period of time thathas passed. Nevertheless, some observations can be made on each case.

The experience in Australia, where investment response has been in theface of robust growth in demand, suggests that governments need tocarefully consider the implications of their policies and subsequentactions on private investors that the reforms are attempting to attract.If investors expect that governments will intervene in wholesale marketsto prevent prices from rising to sufficiently high levels to recover costsof peaking capacity, then attracting that investment will become moredifficult and hence undermine reliability objectives. The experience inSouth Australia suggests that markets can respond to price signals and

14

EXECUTIVE SUMMARY x

meet demand and reliability requirements where government policiesare consistent with the development of efficient and sustainableelectricity markets, and where they are implemented transparently andconsistently.

The Canadian experience with retail price caps appears to be yieldingtwo very different results in terms of investment. The Alberta marketexperience suggests that a sufficiently high cap on prices may not deterinvestment in new capacity. However, the recent experience in Ontarioshows that actions by government to intervene with low price caps candeter investment. Indeed, such political intervention destroys the integrityof the market by destroying incentives to invest and by creating moralhazard. If generators believe that customers will be protected bygovernment intervention from spikes that are needed to recover fixedcosts, then markets will fail to deliver new capacity.

This is not to suggest that price caps may be needed in wholesale marketsat any time. When there is a demonstration of market power beingexercised, as was seen in the California market, price caps may be neededas a temporary instrument to prevent excessive profits from being taken.However, such a measure should be transitional only, and should bephased out as quickly as other mechanisms to address market power,such as enhanced demand response, can be implemented.

By contrast, the New Zealand situation presents quite unusualcircumstances, because the combination of its heavy reliance on hydroresources and its lack of interconnection with other non-hydro systemsmakes it vulnerable to energy shortages that are difficult to predict.Having otherwise surplus capacity available for infrequent dry years doesnot appear to be economic.

The government proposal to contract for reserve capacity that would beset aside from the market and be offered only during dry years wouldaddress this problem.The main difficulty with this proposal is the termsunder which the capacity would be released to the market. Thegovernment has stated that it intends to release the capacity in themarket only during times of shortage. It will, however, be difficult to definethe level of shortage and the quantity to be released in a way that doesnot influence the behaviour of market participants. Disruption would be

15

EXECUTIVE SUMMARY

minimised if the capacity were only to be made available at very highprices. But there will be pressure on the government to release capacityto the market whenever prices rise. In this regard, the situation would beanalogous to those pressures on IEA member governments to release oilstocks to the market.

In Norway, the government faced considerable dissatisfaction with highelectricity prices. However, there were at least three factors that helpedgive the government greater confidence to rely on market mechanisms toresolve the crisis. First, there had been a long history of open electricitymarkets in Norway, a result which had led to a better utilisation ofgenerating capacity and lower electricity prices over several years. Thefact that customers had already enjoyed several years of benefitsincreased confidence that the electricity markets do create benefits forend consumers. Second, the opening of the market internationally,thereby getting access to additional supplies from neighbouring countries,helped to reduce the risk that the market could be manipulated whencapacity was tight. Finally, the existence of the international electricitymarket also meant that effective intervention by government would haveto be co-ordinated with action by the governments of the othercountries served by this market. In fact, such co-ordinated discussionsamong Nordic ministers are carried out on a regular basis. Thismechanism is much better suited to considered joint action rather thana short-term response.

Thus, while electricity markets may be delivering adequate levels ofinvestment, price spikes are testing government commitment to allowmarkets to sort things out. Concentration of electricity markets andconcerns about the manipulation of prices in some markets, such asCalifornia, make it difficult for a government or a regulatory body todetermine if prices are reflecting scarcity or are the result of the exerciseof market power. In some cases, particularly when smaller consumers areexposed to these price spikes, this has sparked government interventionin electricity markets.

Protection of consumers from high prices must be carefully designed toavoid disruption of the market. Intervention by governments in theelectricity markets can create regulatory uncertainty which can

16

EXECUTIVE SUMMARY x

discourage investors. Thus, interventions such as default supply optionsfor small consumers should be chosen carefully, with an awareness of therisks involved. Price caps, if employed, should be set before spikes occur,and at sufficiently high levels. They should also be transitional measuresuntil a more workably competitive market can be established.Governments must also work in co-operation with governments ofinterconnected markets to ensure that measures do not work at crosspurposes.

Addressing the very high price volatility experienced in electricitymarkets is most efficiently done by addressing its cause: the very lowdemand-price elasticity of electricity consumption.There is considerableevidence that this elasticity is lower than it could be because of the lackof ability and incentives for demand to respond to price. Enhancingdemand response will reduce the extreme prices experienced duringtight supply, in effect by widening the price peaks over a larger number ofhours.This will create a more stable environment for power generationinvestment and should increase confidence that electricity markets canbe used to ration capacity by price, ensuring that the supply of electricityremains reliable.

Price volatility also raises questions about the markets ability to ensurea secure and reliable supply. Attracting sufficient investment should notbe a problem in OECD countries since consumers place very high valueon electricity and pay cost-recovering prices.The difficulty is really findinga model that properly values the security of the electricity supply.The oldmonopoly system worked but at the cost of losing economic efficiency.A new system based on bilateral contracts between producers andconsumers can also work, but consumers need to place a value on theirsecurity of supply. For large consumers, this is not a problem, but theissue for small consumers is less clear, since consumers may be less awareof the price risks. Governments have a role in making consumers awareof these risks.

As a consequence, the governments security of supply policy is tied upwith its policies affecting new investment. Some measures can be taken,for example, to remove obstacles to new investment by streamliningapproval of new generating plant.

17

EXECUTIVE SUMMARY

However, mechanisms that intervene directly in electricity markets, suchas capacity market mechanisms, can have much stronger effects on thecost of electricity. Several governments have recently reviewed capacitymechanisms and have rejected their use because they expected suchmechanisms would increase the cost of electricity and they questioned theeffectiveness of these mechanisms at stimulating new investment.Nevertheless, a well-designed capacity mechanism that requires retailersto have arranged adequate resources during peak periods might helpprovide incentives for retailers to acquire sufficient peak capacity (or towork with customers to have sufficient demand response).

The report has the following five recommendations:

Define clearly the governments role in electricity marketreform and the terms of its involvement as precisely aspossible. Attracting investment in power generation requires a clearmarket design, with predictable changes and no interference intothe market or into the operation of the independent institutionsestablished to implement the market reform. The governments rolemust be clearly set out both as the agent of the reforms and in itsenergy policy involvement once the market opens.

Recognise that electricity price fluctuations are intrinsic towell-functioning electricity markets. Allowing markets to signalthe need for new investment in generation means that prices willgo high on occasion. Governments need to anticipate that suchfluctuations will occur and ensure that consumers are aware of pricerisks and have options to mitigate these risks.

Develop demand response within electricity markets.Fluctuating spot electricity prices offer rewards as well as risks.The lowprice elasticity of electricity demand, especially for small customers, isat least partly due to the inability to reward consumers for adjustingtheir consumption when prices are high. Greater demand response inelectricity markets is needed to help ensure that electricity marketsare always able to clear, i.e. by rationing electricity supply according toprice rather than through brownouts or blackouts. A stronger demandresponse will help mitigate market power in electricity markets andprovide potential investors with more predictable energy (and ancillaryservice) prices and therefore decrease investment risks.

18

EXECUTIVE SUMMARY x

Be prepared to detect and to act upon wholesale electricitymarket manipulation. In order to address concerns aboutwholesale electricity market manipulation, governments must ensurethat electricity markets have monitoring mechanisms that can not onlydetect manipulation as it is occurring but also take prompt action tomitigate its impacts.This will reduce pressure on the government torespond e.g. through direct price caps which could drive away neededinvestment.

Monitoring adequacy of gas markets and investments. Thepreference of investors in some markets for CCGT for building newpower generating capacity means that gas markets assume a greaterimportance than ever for power generation development. Forgovernments, this means moving forward on liberalisation of both thegas market and the electricity market, and monitoring the adequacy ofinvestment in both gas and electric infrastructure.

19

EXECUTIVE SUMMARY

INTRODUCTIONElectricity markets are opening to competition in nearly all IEA countries.Most IEA member governments have implemented laws to liberalisewholesale electricity markets and allow at least some of the largeconsumers the opportunity to choose suppliers, to permit electricitygenerators and consumers to have non-discriminatory access totransmission and distribution systems, to liberalise electricity trading ona bilateral basis and on organised exchanges, and to allow for free entryof new producers, on non-discriminatory conditions.

The main objective of electricity market liberalisation is to improve theeconomic efficiency of the electricity supply industry.The introduction ofcompetition in the generation and retail supply of electricity is expectedto improve productive efficiency by reducing operating costs andimproving allocative efficiency by aligning prices with costs.

There is a great deal of evidence to suggest that market liberalisation has,in general, led to a reduction in operating costs of generating plants byimproving labour productivity, reducing maintenance costs and improvingfuel purchasing strategies3. As a number of electricity markets have beenopened in the presence of surplus generating capacity, there have beenvery strong competitive pressures to reduce such costs as prices havefallen. Indeed, all evidence indicates that such cost reductions areoccurring, and that firms are becoming more efficient in response to pricesignals.

As a capital-intensive industry, a key sign of an efficiently functioningelectricity supply industry is an efficient allocation of capital. Previous IEAwork4, Security of Supply in Electricity Markets: Evidence and Policy Issues,examined the question of the adequacy of the investment levels in sevenreformed markets. Its main conclusions are:

Substantial investment has taken place since market liberalisation andOECD electricity markets are generally reliable, the exception ofCalifornia notwithstanding.

21

INTRODUCTION1

3. See, for example, IEA, 1999.

4. IEA, 2002a.

Reserve margins have fallen generally, consistent with theimprovement of allocative efficiency.

New capacity investment favoured the most economic option: naturalgas where this was available but also coal where this option was lessexpensive.

It was too early to conclude whether electricity generationinvestment would mimic boom and bust cycles observed in otherindustries.

Markets may increase flexibility of the demand side (e.g. through load-shifting or distributed generation) which would reduce the size of thereserve capacity required.

The biggest challenges remain ahead most markets are just beginningto approach their first major investment cycle5.

However, the overall level of investment is but one of the major issuessurrounding power generation investment in liberalised markets. Arecent IEA study has looked at the level of investment required and howthis investment will be attracted to the power market. It concludes thatthe power sector of OECD countries will need $4 trillion of investmentbetween 2000 and 2030. About half this amount would be needed forpower generation investment6.

This report will focus on three other issues. The first is the impact ofliberalisation on power generation investors and on the technologieschosen for investment. It will consider how investors have adapted to themarket in terms of their analytical methods and their corporatestructures in response to the greatly increased risks that need to beborne. It will then consider the consequences of these changes for thechoice of power generation technology for investment.

Investment in global power generation has been growing strongly.Figure 1 shows the global orders for new power plants since the 1950s,with orders for gas turbines shown separately starting in 1991. It shows

22

5. IEA, 2002a also examined the question of transmission investment and concluded that there was a need for policies toencourage investment in transmission in many OECD regions, a message reinforced by the blackout in the United States andCanada in August 2003.The present report is focused strictly on investment in power generation.

6. IEA, 2003a.

INTRODUCTION x1

a steady growth trend with two boom periods one from the mid-1960suntil the early 1980s and a much shorter boom in the late 1990s thatpeaked in 2001.

The figure also shows the emerging role that natural gas-fired turbineshave come to play in power generation, accounting for 65% of thecapacity ordered since 1991. While this emergence is partly due toimprovements in gas-fired generation technology, there is no questionthat in many liberalised markets gas-fired generation has beenfavoured.

This preference for natural gas-fired generation is expected to continueinto the future.The 2002 World Energy Outlook (WEO) suggests that, givencurrent policies, over half of the 2 000 GW of generating capacity addedbetween 1999 and 2030 in the OECD will use natural gas (Figure 2).

The strong preference for gas-fired power generation has raisedconcerns about energy diversification. The reason for this is that, as

23

INTRODUCTION1

Figure 1

Global Power Generation Orders, Including Gas Turbines (1950-2003)

Source: Siemens Power Generation (2003 is estimated).

0

20

40

60

80

100

120

140

160

180

200

Year

Gen

erat

ion

Ord

ers

(GW

)

Total Gas Turbines

195019

5219

5419

5619

5819

6019

6219

6419

6619

6819

7019

7219

7419

7619

7819

8019

8219

8419

8619

8819

9019

9219

9419

9619

9820

0020

02

shown in the World Energy Outlook 2002, the increased demand forelectric power is the main driver for investment in all commercialprimary energies but oil. Therefore, the expectation that powergeneration markets in the OECD will be increasingly attracted to naturalgas will have a real influence on the primary energy mix and on thediversity of electricity supply, more than doubling its share to 32% of the2030 fuel mix.

The issue to be addressed is whether these investments are consistentwith economic efficiency or whether short-termism on the part ofinvestors in liberalised markets is leading to decisions that areinconsistent with an economically rational allocation of capital. Chapter 2will examine how investors have adapted to reformed electricity marketsthrough the need to internalise various investment risks in their decision-

24

INTRODUCTION x1

Figure 2

Share of Projected Capacity Additions in OECD GeneratingCapacity by Fuel (1999-2030)

Note: Fuel cells are accounted separately but are assumed to use natural gas.Source: IEA, 2002b.

Coal20%

Oil4%

Gas48%

Nuclear4%

Hydro (incl. Pumped Storage)

4%

Other Renewables15%

Fuel Cells5%

making. It discusses the nature of the risks that now need to beinternalised and compares qualitatively the risk attributes of differentpower generation technologies. It will discuss the new techniques forquantification of such risks, and the implications of the results of suchanalyses for investors. It will examine the consequences on the choice oftechnologies for power generation.

The second issue is how power generation investors are adapting toinvestment risks. At the time of market liberalisation, it was argued thatthe development of financial instruments would help investors byproviding financial hedges for future electricity or gas prices. In fact, suchmarkets have been slow to develop and appear unlikely to offer long-termfinancial hedges that might offset price and fuel cost risks. As aconsequence, firms are seeking to hedge these risks by the use of long-term contracts, by integration of generation with retail, by growing in size,and by mergers between gas and electricity companies. Chapter 3 willlook at these strategies and discuss whether such structural changes areconsistent with efficient energy markets.

The third major issue to be addressed in the study is inspired by recentdevelopments in several OECD electricity markets. During the past twoyears, markets in Australia, New Zealand, Canada and Scandinavia have allexperienced electricity price spikes.The tendency of deregulated powermarkets to produce price spikes can depending on market design andconsumers choices significantly raise bills for small consumers.This hasin turn created what can be termed an electricity price crisis andpushed the issue to the top of the agenda of the governments concerned.Table 1 summarises the major crises in several jurisdictions.

The causes of crises are varied. Unlike California, market manipulation orbad market design does not appear to have played a significant role in theother cases. Like California, governments in all these markets have beenplaced under considerable political pressure to respond to the concernsof consumers about prices and the security of supply.

These electricity price crises and the governments responses to themare critical to understand the future of using spot prices in electricitymarkets to signal the need for new investment. Chapter 4 will briefly

25

INTRODUCTION1

examine price crises in Canada, Australia, New Zealand and Norway, anddiscuss how the governments responses will affect future investment.

Chapter 5 will summarise the main conclusions of the report and offerpolicy recommendations.

26

INTRODUCTION x1

Jurisdiction Cause of Price Crisis Duration

California Tight capacity (low rainfall), 5/2000 5/2001manipulated markets

Canada High fuel prices, high electricity 9/2000 3/2001(Alberta) prices in neighbouring markets

New Zealand Tight energy supply due to low rainfall 4/2001 7/2001and concerns about availability 4/2003 6/2003

of thermal fuels

Australia Tight capacity due to rapid growth 1/2002 -3/2002(South Australia)

Nord Pool Tight energy supply due to low 12/2002 03/2003rainfall/cold winter

Canada Demand growth, delay in capacity 7/2002 7/2003(Ontario) investment

Table 1

Summary of Markets Experiencing Electricity Price Crises

RISK AND POWER GENERATIONINVESTMENT

Introduction

Prior to the liberalisation of energy markets, energy firms were able tooperate as integrated monopolies able to pass on all costs of investmentsto energy consumers. For example, in the electric power sector, utilitiescould expect the cost of their prudently incurred investments in powergeneration, including an adequate rate of return, to be recovered fromconsumers. In view of that guaranteed rate of return, utilities couldfinance their investment with a low share of equity and borrow atinterest rates close to government debt yields.There was no market risk.The main risk was the risk of unfavourable regulatory decisions and costoverruns due to bad project management. Overinvestment could beaccommodated as excess reserve margin since excess capacity did notcreate a reduction in electricity prices.

In such an environment, most of the risks associated with suchinvestments were not directly a concern of the energy company.Increased costs, if demonstrated to be prudently incurred, could bepassed on as increased prices.There was little incentive for companies totake account of such risks when making investment decisions.

The introduction of liberalisation in energy markets is removing theregulatory risk shield and exposing investors to various risks or exposingthem to risks in different ways. Generators are no longer guaranteed theability to recover all costs from power consumers. Nor is the future pricelevel guaranteed.

The issue of interest here is how the internalisation of risks affectsdecision-making on investment.This chapter discusses new techniques forquantification of such risks, and the implications of the results of suchanalyses for investors. It will examine the consequences on the choice oftechnologies for power generation, and how the structure of firms ischanging in response to perceived risk.

27

RISK AND POWER GENERATION INVESTMENT2

Power Generation Investment Risks in the Liberalised Market

Investment in power generation comprises a large and diverse set ofrisks. Business risks include:

Economy-wide factors that affect the demand for electricity or theavailability of labour and capital.

Factors under the control of the policy-makers, such as regulatory(economic and non-economic) and political risks, with possibleimplications for costs, financing conditions and on earnings.

Factors under the control of the company, such as the size anddiversity of its investment programme, the choice and diversity ofgeneration technologies, control of costs during construction andoperation.

The price and volume risks in the electricity market.

Fuel price and, to a lesser extent, availability risks.

Financial risks arising from the financing of investment. They can tosome extent be mitigated by the capital structure of the company7.

The level of risk anticipated by an investor in a power plant will bereflected in the level of return expected on that investment.The greaterthe business and financial risks, the higher the return that will bedemanded.

The most fundamental change affecting the value of investments inliberalised markets is the inherent uncertainty about electricity prices inelectricity markets.The uncertain future level of prices from investmentin generation creates a risk for the investor. While this risk affects allgenerating technologies8, it does so in different ways.Technologies whichhave a higher specific investment for capacity even though they may haverelatively low fuel costs (wind, nuclear) are more greatly affected by thisrisk because there is less they can do to respond. Thus, although highcapital cost and low fuel cost technologies will likely be competitive in the

7. IEA, 1994.

8. Conversely, if e.g. renewable energy technologies are favoured by long-term contracts at fixed prices, this significantlyreduces this risk compared to technologies that would need to take an uncertain market price.

28

RISK AND POWER GENERATION INVESTMENT x2

short-run and therefore produce electricity, they will be more exposedto cover capital employed. A firm reliant on such technologies may finditself in financial difficulties if prices slump for a prolonged period. Therecent experience of British Energy in the UK electricity market leadingto action by the UK government to support the company is a case inpoint9.

Uncertain electricity prices also expose projects that have a long lead andconstruction time to additional risks. Economies of scale favour largepower projects over small ones as capital costs per kW for a giventechnology generally decrease with increasing scale, or at least appear todo so. However, the combination of a long lead time, uncertain growth indemand for electricity and price, and uncertainty in the total cost offinancing construction increase risks for larger projects. Furthermore,very large projects that must effectively be built as a single large plant (e.g.a very large hydro dam) are more vulnerable to this type of risk thanprojects for which development can be phased in as several smallerpower plants in response to market conditions.

The cost of fuel can be a significant additional risk to profits, particularlyfor technologies where fuel costs are a high proportion of totalgenerating costs. Natural gas technologies are particularly sensitive tofuel prices and price volatility, as fuel costs tend to constitute the majorityof generating costs.

Uncertainty in future natural gas prices is increased by the liberalisationof the natural gas market, and the disappearance of long-term contractsavailable for the supply of natural gas for power generation. High volatilityof natural gas prices also will tend to increase short-term risks associatedwith natural gas. If rises in natural gas prices accompany falls in electricityprices, and the generator has not financed the project in recognition ofthis risk, the financial distress for natural gas power generators can besevere. A number of cogeneration facilities in the Netherlands foundthemselves in precisely this situation leading to action by theNetherlands government to support cogeneration10.

29


9. Hewitt, 2002.

10. IEA, 2002c, p. 63.

The situation for investors is further complicated by the outlook forresource availability and cost. Although sufficient economic resourcesexist for natural gas and other fuels used in power generation11, sufficientinvestment is needed in infrastructure to produce and transport fuels topower plants. A further consideration is the source of the future gassupplies.The source of natural gas supplies is expected to shift stronglyover the next 30 years, resulting in a quadrupling of OECD natural gasimport volumes, and increased reliance on production from outside theOECD12. While resource rents from the development of existingdomestic natural gas resources or from gas imports from IEA countrieshave been relatively predictable, rent-seeking by non-IEA gas-producingcountries may be a significant long-term uncertainty.

The key question for an investor in fossil-fired power generation in anopen market will be the level and development of the difference betweenthe price of electricity and the cost of fuel used to produce it the so-called spark spread.The importance of the spark spread will depend onthe type of power plant and how it is intended to be used. For baseloadpower plants, a relatively large favourable spark spread is desirable so thatthe plant can operate at all times to recover the specifically higher capitalcosts of such a plant over a large number of hours. For peaking loadplants, with higher fuel costs, capital costs must be recouped over asmaller number of hours.Thus, peaking plant characteristics will favour aflexible generating plant that is able to take advantage whenever the sparkspread is favourable.This requires not only technical flexibility to respondto changing prices, but, in the case of natural gas, flexibility in starting andstopping gas offtakes in line with the spark spread.

The market rules themselves can be a source of risk and will affectdifferent technologies differently. For example, changes in electricitymarket rules that put an opportunity cost on unreliability of outputhave affected the cost of wind power in the United Kingdom. Similarly,price signals that encourage more efficient use of the electricity gridwill also favour technologies that can locate to take advantage of theseincentives.

30

11. IEA, 2001.

12. IEA, 2002b.


The costs of additional emissions controls, formerly passed directly on toconsumers, must now also be considered as a risk to the profitability ofpower investments. Existing emissions controls on coal-fired plantsinclude particulates and sulphur dioxide. Both coal and natural gas plantsare covered by emissions controls on nitrogen oxides. However, futureregulatory actions to lower allowable levels, or to introduce controls onother substances such as mercury, remain a risk for investors. Nuclearpower plants are restricted in their emissions of radionuclides and maybe subject to additional safety regulations.

Probably the greatest uncertainty for investors in new power plants willbe controls on future carbon dioxide emissions. In the European Union,an emissions trading directive will be issued in 2003. Canada has alsoindicated that they will use emissions trading to control emissions fromlarge stationary sources of carbon dioxide (CO2).The unknown value ofcarbon emissions permits and the mechanism chosen to allocate permitswill become a very large and potentially critical uncertainty in powergeneration investment. This uncertainty will grow in the future,particularly as future restrictions on levels of carbon dioxide emissionsbeyond the first commitment period of the Kyoto Protocol are unknown.For investment in fossil-fired generation, the price of permits will directlyaffect the profitability of power plants. The price of the permit is alsoexpected to have a strong influence on the price of electricity and willfurther increase uncertainty about future electricity prices.

The risks associated with gaining approval to construct a new powerplant also differ by technology.The risk is lower and the time span for theapproval process is usually shortest for gas-fired power plants and smallpower plants such as fuel cells and photovoltaics. Although this riskalready existed in regulated markets, the ability to pass through theapproval costs to consumers is no longer automatic.

The important point for power generation is that the nature of the risks(the risk profile) is different for different types of generation technologyand fuels (Table 2). Thus, even when levelised costs are equivalent andtechnologies are commercially proven, different risk profiles of differenttechnologies can influence the choice of power generation mix, the rangeof technologies offered, and the strategies for their development andoperation.

31


Gas-fired technologies have characteristics that should be favourableunder these conditions. The relatively low capital cost, short lead time,standardised design and, for some technologies, flexibility in operationprovide significant advantages to investors. On the other hand, natural gasprice uncertainty remains a large risk to the investor.

Nuclear power plants, by contrast, have a relatively low proportion of fueland operating costs but high capital cost. Furthermore, economies ofscale have tended to favour very large plants (1 000 MW and above)resulting in a relatively large capital commitment to a single constructionproject and hence associated investment risk. Newer designs are moreflexible with regard to operations.The potential economic advantages ofbuilding smaller, more modular nuclear plants are also being explored bysome nuclear power plant designers.

Coal power projects have also tended to become more capital-intensiveto take advantage of economies of scale, to meet tighter environmentalstandards more economically, and to improve fuel efficiency. As withnuclear plants, lead and construction times for coal-fired power plantscan be long.

32


Technology Unit Lead Capital Operating Fuel Co2 RegulatorySize Time Cost/kW Cost Cost Emissions Risk

CCGT Medium Short Low Low High Medium Low

Coal Large Long High Medium Medium High High

Nuclear Very large Long High Medium Low Nil High

Hydro Very large Long Very high Very low Nil Nil High

Wind Small Short High Very low Nil Nil Medium

Recip. Engine Small Very short Low Low High Medium Medium

Fuel Cells Small Very short Very high Medium High Medium Low

Photovoltaics Very small Very short Very high Very low Nil Nil Low

Notes: distributed generation technologies are shaded. CO2 emissions refer to emissions during combustion/reformation only.

Table 2

Qualitative Comparison of Generating Technology by RiskCharacteristics

Hydro projects come in all scales. Larger ones demand substantial leadtime and are exposed to considerable risks during the construction phaseas the length of the project can be subject to delays, the cost ofborrowing can also change and increase the cost of the project. Usuallythere is no borrowing available in accordance with the amortisation timeof a hydro project.The large economic potential of hydro has not beenfully developed in some developing countries because of the verysubstantial risk premium resulting from the high sovereign risk.Operations by contrast can be highly flexible and able to take advantageof market conditions to optimise profitability. Long-term shifts andvariations in rainfall patterns, e.g. due to climate change, remain anotherrisk factor.

Of other renewables, solar and wind capacity are also capital-intensive.These plants have some very attractive low-risk characteristics, includingvery short lead times, no fuel costs or emissions, and low operating costs(hence little effect should these costs escalate). However, the variabilityof output of wind power reduces the value of the power produced. Onthe other hand, the fact that solar electricity from photovoltaics isproduced at the point of consumption and during daylight hours, whenprices are generally higher, increases the value of its electricity.

Reciprocating engines and fuel cells are two distributed generationtechnologies that use fossil fuels. Like photovoltaics, these distributedgeneration technologies have very short lead times and can be installeddirectly at the site of an electricity consumer. Their flexibility of usemeans that they can be operated during hours when power prices arefavourable.

The competitiveness of reciprocating engines is thus dependent on thecost of delivered electricity that the distributed technology replaces.When fuel prices are low and electricity prices high, owners of thesedistributed generation technologies can produce their own electricityand reduce purchases from the electricity market. Electricity consumersthat own generation facilities thus have a kind of physical hedge on themarketplace that allows them to cap the prices they pay. Reciprocatingengines are also portable, i.e. they can be sold and moved to anotherlocation.

33


Quantifying Investment Risks in Power Generation

While identifying investment risks in power generation may bestraightforward, investors in power generation attempt to understand therelative importance of different risks by quantifying them where possible.For the more important risks, it will be prudent to adopt risk managementstrategies that can cost-effectively reduce exposure to such risks.

Thus, estimating the profitability of an investment must rely on modelling,requiring knowledge of the future costs and the future revenues of thegenerating project and their variation. For a company with choice amonggenerating options, with different lead times, different uncertainties incosts, a uniform methodology must be developed to comparetechnologies with different characteristics.

The levelised cost methodology, a widely accepted costing method forinvestments, has been a useful tool for investors and for overall economicanalysis because it evaluated costs and energy production and discountedthem to take account of the time value of money. It provided an objectivebasis on which to provide a comparison of different technologies, e.g. forbaseload power generation.This approach has been used in the NEA/IEAstudy, Projected Costs of Generating Electricity, last updated in 199813. Thisapproach reflected the reality of long-term financing, passing on costs to the(captive) customers, known technology paradigms, a predictable place in themerit order, a strong increase in consumption and a short build-up time forselling the output of a new plant.The levelised cost approach involves:

Developing estimates for all major cost components for a selection ofgenerating technologies comprising:

Capital costs, including initial expenditures (as well as associatedinterest costs during the construction period), ongoing capital costswhile the plant is operational, and decommissioning costs.

Operating and maintenance costs.

Fuel costs based on forecasts of fuel prices and fuel conversionefficiencies for the generating technology (and fuel waste disposalcosts as applicable).

34

13. NEA/IEA, 1998.


Estimating the average annual energy production from the powerplant according to assumptions about technical availability.

Discounting the stream of costs to estimate their present valueaccording to an assumed discount rate (5% and 10% were used).

Using the same discounting procedure to estimate the present valueof the energy production.

Taking the ratio of the costs and energy output to obtain a levelisedcost of power production.

This discounted cash flow (DCF) approach remains useful for providing acomparison among power generation technologies. Power companies willapply DCF methodology based on an internal target for return on equity(the hurdle rate) to make a decision whether to invest or not and todecide between different projects. To assess various risks, differentscenarios or sensitivities will be calculated, which often give a goodassessment of the risks involved.

However, in an electricity market, what matters to the investor is theprofitability of the investment against the risk to the capital employed.Provided that the market is operating efficiently, the investors will makethe choice of a generating technology that incorporates risks and is alsothe most economic choice available. Unfortunately, it is difficult for thelevelised cost methodology to incorporate risks effectively.Thus it needsto be complemented by approaches that account for risks in future costsand revenues.

New financial techniques are becoming available that help quantify theimpact of these risks on the costs of different options. Such assessmentscan help investors make better decisions. Investments with lower riskshould have correspondingly lower hurdle rates for investment.

For example, a recent study carried out at the Massachusetts Institute ofTechnology14 on the future of nuclear power presented an economiccomparison of generating technologies using a methodology thataccounted for risk. In its approach, higher returns on equity and a highershare of relatively costly equity in the capital investment were assumed

35


14. MIT, 2003.

for nuclear power investment compared to coal or natural gas. Theanalysis also included the effect of taxes, which also increases the effectivereturn on equity required for all technologies.

The results of this analysis are summarised in Table 3. The base scenariofound that nuclear power was more expensive than coal or natural gas bya very wide margin (USD 25 per MWh). Using the same input assumptions,one can also calculate levelised unit energy costs using the method in theNEA/IEA study, i.e. which applies the same discount rates to all technologies.In this case, the gap between nuclear and CCGT is much smaller, even at10% real discount rate. A significant part of the difference between theestimates can be attributed to the higher perceived risk and hence higherhurdle rate employed for nuclear power rather than other technologies.

Another simple technique is to use a probabilistic assessment to look at awide range of uncertainties in key risks, e.g. natural gas costs and electricityprices. The resulting distribution of outputs gives both an expected valueand a range of probabilities that an investment would be profitable. Theresults of one such simulation, performed for the US Energy InformationAdministration, are illustrated in Figure 3 for an investment in a natural gascombined cycle plant15.An analysis that assumes certain fuel and electricityprices yields a positive net present value. By contrast, a probabilisticassessment shows that the investment would stand about an 83% chance of

36

15. EIA, 2002.


Technology MIT Study Levelised Cost at 5% Levelised Cost at 10%Base Case Discount Rate Discount Rate

Nuclear 67 44 55

CCGT 41 44 45

Coal 42 33 40

Sources: MIT, 2003 and Secretariat analysis.

Table 3

Comparison of Levelised Unit Energy Cost Estimates UsingDifferent Methodologies (USD/MWh)

being profitable over a 20-year period, given the forecast prices forelectricity and natural gas and their uncertainty.

A second approach that can improve a discounted cash flow analysis is todiscount costs and revenues at different rates according to their riskiness.Thus, for a power plant with more or less fixed costs but very uncertainrevenues the costs can be discounted at a low rate but revenues at avery high rate reflecting the uncertainty in future prices the result canbe an even higher effective discount rate. If a major cost component (suchas fuel cost) is highly uncertain, its costs should be discounted at an evenlower rate (the present value of costs is increased because ofuncertainty).

Real Options and Investment in Peaking Plants

Estimating the profitability of peaking plants is a particularly difficultexercise in liberalised power markets. Peaking plants, by definition,depend on the high prices that appear over a relatively small number of

37


Figure 3

Net Present Value (NPV) Frequency Range for CCGT Investment

Source: Energy Information Administration.

hours to earn a return on the initial investment. The volatility ofelectricity and fuel prices makes prediction of the actual profits moredifficult16.

The fact that both fuel and power prices can be volatile can create notonly risks, but also opportunities for the owners of peaking plants tomake profits, producing electricity whenever the spark spread thedifference between the electricity price and the fuel cost is favourable.

Discounted cash flow analyses tend to rely on single estimates of fuel andpower prices, and thus underestimate profits that can be earned bypeaking plants.A more accurate method needs to estimate how likely in any hour the favourable spark spread will occur. Sophisticatedanalytical approaches applying financial options methods can provide ameans to estimate how many hours such plants are expected to operate,and the profits that can be expected from their operation.

A technique borrowed from financial options theory, known as realoptions, can be used to develop estimates of future uncertainty in pricesgiven the observed volatility. By analysing the volatility of forward pricesof electricity and key cost components (e.g. fuel costs), an estimate of thelong-term uncertainty in profitability can be obtained17. It is this impacton long-term profitability than can influence investment.

Incorporation of uncertainty of fuel costs into estimates of profitabilitywould appear to show the increasing importance of fuel cost uncertaintyand volatility for profitability. However, these estimates must take accountof at least two factors. In the short term, electricity and natural gas pricesare sensitive to the supply and demand balance and have proven to bevery volatile, a fact reflecting, among others, the relatively low short-termelasticity of demand for these commodities. In the long term, however,price movements are dictated by demand and supply fundamentals e.g.when prices are high enough, new generating capacity is added (or new

38

16. In considering the role of price risk in investment, it is useful to take account of the difference between the volatility offuel or electricity prices and the uncertainty in these prices.Volatility is a measure of the degree of change in prices from onetime period to the next. Electric power prices are highly volatile: prices can change from hour to hour by a factor of two ormore and depend strongly on very specific factors such as the demand-supply balance or the price elasticity of demand.Volatility can be regarded as a measure of the rate of change of uncertainty in the price.

17. Geman, 2003.


gas wells are drilled) to pull prices back towards long-run marginal costsof new generating capacity.This phenomenon of mean reversion meansthat power prices and costs of major components of those prices suchas fuel costs are less uncertain than the short-term swings in theseprices might imply.

The second factor is the extent to which natural gas prices are positivelycorrelated with electricity prices. If an investor in a natural gas powerplant can be assured that electricity prices will increase whenever naturalgas prices increase (as a consequence of natural gas-fuelled generatorsbeing the marginal source of electricity), then the impact of natural gasprices on generation prices would be quite modest. Indeed, the EIA studycited above18 assumed a positive correlation coefficient betweenelectricity and gas prices of 0.88, based on its analysis of Henry Hubnatural gas spot prices versus an average of electricity spot prices from1999 to 2002.

The role of volatility can also be important depending on the type ofplant. For the purposes of investment in baseload power generation,volatility is relatively unimportant compared to the expected averageprice level, as the plant is intended to be operating in nearly all hours. Bycontrast, a peaking plant will only be profitable to operate in a relativelysmall number of hours depending on a high enough spark spread in thosehours. Fuel and power price volatilities create opportunities for theowners of peaking plants to make profits.

It is difficult to estimate the value of a peaking plant using discounted cashflow methods because these techniques tend to rely on single estimatesof fuel and power prices. The very high uncertainty in price leads todiscounting at very high rates and can lead to the conclusion that theseplants are unprofitable. By relying on average conditions, however, thesetechniques underestimate the profit possibilities of those peaking plantsto be employed only when the price of electricity is greater than the costof producing it and when the plant is able to sell the gas if gas prices arehigh and electricity prices are relatively low. In other words, the value ofthe flexibility of a peaking plant tends to be underestimated.

39


18. EIA, 2002.

The real options method is a technique that can be used to estimate thevalue of a peaking plant, by valuing its ability to take advantage of volatilefuel and electricity prices19. In this method, the power plant itself istreated as a financial option, except that it is a real physical asset apower plant rather than a financial security. The value of the powerplant then becomes a series of call options for each hour during whichthe power price or the fuel price can vary.The implicit assumption is thatthe power plant will operate in each and every hour in which it isprofitable to do so, and will not operate in those hours in which it is not.Knowing the volatility in power and fuel prices helps quantify theuncertainty and the likelihood of profitability in any future hour.

Analysis of investment in a coal-fired mid-merit plant and a gas-firedpeaking unit by using this technique suggests that both units havesignificant flexibility value that would actually be greater than thevaluation suggested by the discounted cash flow approach. Results fromthis type of analysis suggest that prices paid for mid and peak powerplants in the United States in the late 1990s can be better explained byreal options approaches than by conventional discounted cash flowmethods20.

Value of Flexibility in Power Plant Development

Traditional investment analyses place limited emphasis on the timing ofthe investment versus market conditions. In these approaches, the valueof developing a single large plant is considered rather than a series ofsmaller plants, although the smaller plant strategy might have anadvantage, e.g. if the growth in demand for electricity turned out to belower than expected. Without a method to quantify the value of thisflexibility, the economies of scale evident in larger power plants (in termsof installed cost per kW) trump any concern about risks of addingcapacity in larger units.

In energy markets, when future prices are uncertain, investors are awareof the potential value of proceeding incrementally in developing new

40

19. White and Poats, 2000.

20. Frayer and Uludere, 2001.


capacity. Quantitative methodologies that assess the value of being ableto defer the decision of making part of the investment until marketconditions become clearer have therefore been developed. Indeed, one ofthe early applications of the real options approach was to assess the valueof phasing the development of oil fields versus a strategy of developing allat once21.

Research to apply the same techniques to assess the flexibility value ofdifferent strategies for power generation project development is still atan early stage. One study has applied the real options method to estimatethe value of developing wind power projects in stages, rather than all atonce, in recognition of the inherent flexibility of smaller plants22.Anotherstudy examined the cost-effectiveness of developing an IGCC plant inphases, initially using natural gas as a fuel and thereby delaying thedecision as to when to convert the plant to using gasified coal23.

Despite these interesting academic results, the real options approach hasachieved little acceptance by power generation investors to date.Calculating the real options value of a power plant has proven to be a lessreliable indicator of value than financial options are in the stock market,for a variety of reasons. Unlike financial markets, forward markets forelectricity and natural gas prices are not sufficiently liquid. The modelsmust therefore rely on forecasts of future electricity and fuel prices.These forecasts, and the correlation between electricity and natural gasprices, are highly uncertain in light of changing volatility of these prices.

Summary

The reform of electricity and gas markets has led to major changes in theway decisions are taken on power sector investment. Opening the sectorto competition has led to the internalisation of risk in investmentdecision-making. Investors now examine power generation optionsaccording to the different financial risks posed by the differenttechnologies.

41


21. McCormack and Sick, 2001.

22. Venetsanos et al., 2002.

23. Smeers et al., 2001.

Given the long-term nature of electricity investments, investmentdecisions in baseload generating capacity are being made on the basis oflong-term fundamentals rather than looking at short-term behaviour inthe spot or forward electricity markets. Conventional discounted cashflow methods are still most often used. Nevertheless, investors arebeginning to take account of differences in risk levels in assessing thelikely profitability of different investments.

The current market preference for gas-fired power generation forbaseload power generation in many OECD countries can be explainedmainly by the perceived lower cost of gas-fired generation. Thecharacteristics of the CCGT, its low capital cost, and its flexibility haveadded to its attractiveness. The importance of CCGT means that gasmarkets assume a greater importance than ever for power generationdevelopment. For governments, this means moving forward onliberalisation, and monitoring investment in both gas and electricityinfrastructure.

The preference for gas-fired power generation does expose investors toincreased fuel price risk.The creation or development of electricity andnatural gas markets has led to a system where, in the absence of hedgingpossibilities, the risks of price development can no longer be managed,but must be assessed by probabilistic approaches.

The adequacy of investment in peaking generation remains an even moresensitive issue.The low numbers of hours of operation of peaking plantshave led some to question whether markets can bring forward adequatepeaking capacity. Low capital and high flexibility in operation areparticularly important attributes in an attempt to value peaking capacity.While investors have long been aware of the qualitative benefits offlexibility in a liberalised market, new techniques have now beendeveloped to quantify the value of flexibility. In particular, this researchsuggests that flexible peaking capacity may be much more profitable thantraditional approaches would predict.

42


HEDGING RISK IN POWER GENERATION INVESTMENT

Introduction

Prior to liberalisation, investment in the power sector was a relatively lowbusiness risk and, in many cases, state ownership made access to debtcapital relatively easy. Even for independent power producers, theavailability of a long-term contract, which would pass the marketing riskthrough to a single buyer, made it possible to finance investment at a lowrisk premium.

The liberalisation of gas and electricity markets has changed the natureof corporations responsible for investment. Corporations are no longerdirectly responsible for ensuring adequate supply beyond theircontractual obligations. As always, they remain responsible to theirshareholders and they will invest only if their shareholders think it isprofitable. Many state-owned companies have been privatised. Long-termpurchase contracts for independent power producers are no longercommon. Returns on investment are much more uncertain.The structureof electric utilities has also changed. Some countries have required firmsto legally separate the network activities from the competitive and riskierbusinesses of generation and retail.

The role of the customer is also changing. Customers can choose thelevel of risk they wish to take with regard to the volatility of prices.Theymay also decide to invest in their own on-site power generation with thepotential to sell surplus to other consumers. However, the ability ofcustomers to react to price signals will depend on the type of customer.Furthermore, market opening in many jurisdictions is not yet open tohousehold consumers.

Market liberalisation has also led to the opening of spot electricitymarkets in which producers could sell their output. The emergence ofthese markets encouraged the development of a new type of power plantinvestment the so-called merchant power plant. A merchant powerplant is a plant built where the output is sold at unregulated prices, either

43

HEDGING RISK IN POWER GENERATION INVESTMENT3

into electricity markets or through bilateral contracts. In the late 1990s,during a boom in power plant construction, finance was relatively easy tofind for merchant power plants in US markets. However, thanks to a shiftin market fundamentals (a slowing economy and rising natural gas prices)as well as particular events (Californias electricity crisis, the bankruptcyof Enron), power prices and spark spreads fell dramatically, reducing theprofitability of the new (mostly gas-fired) power plants.The aggregate lossin equity values for companies with merchant plant investments wasmeasured in the hundreds of billions of dollars and a number of projectswere cancelled24.The cost of capital for new plants in the United Stateshas increased as the liquidity of electricity forward markets has greatlydecreased25. As a consequence, it is currently difficult to obtain bankfinance for new merchant power plant construction in the United States.Merchant power plant investment continues to occur in otherjurisdictions, notably in Australia (see Chapter 4).

The risks associated with the merchant plant model have persuaded mostinvestors to find mechanisms to hedge these risks. Financial hedging canuse markets such as forward or futures markets to manage price risk.Contracts between producers and retailers or directly with consumersare another strategy that links the producer and consumer more directly.Finally, there are organisational strategies, mainly mergers, which are nowemerging to deal with the risks associated with new investment or torespond to spark spread risk. Consumers may also hedge their risks bydeveloping their own power plants.

Market Hedges

Financial hedging instruments such as futures and forward markets areimportant tools in the development of efficient electricity markets.Perhaps the best-developed electricity forward and futures markets existin Nord Pool.The futures market Eltermin and options market (Eloption)have developed quite strongly in recent years, with trading volumes nowsignificantly larger than the physical deliveries of 150 TWh/year. Elterminprovides financial contracts for power up to three years ahead. Figure 4

44

24. de Luze,2003.

25. Joskow, 2003.

HEDGING RISK IN POWER GENERATION INVESTMENT x3

illustrates how electricity trade has developed in the Elspot market since1998.Trade through Elspot has actually grown each year and now averagetrade is close to ten times physical deliveries. Eltermin and over-the-counter transactions have also grown (Figure 5) but fell off dramaticallythis past winter, a development attributable to the high electricity prices.

As is apparent from the figures, electricity trade in Nord Pool has takenseveral years to develop. Electricity market trade has also been growingin the United Kingdom and other parts of Europe. However, to dateliquidity is still below a benchmark volume of 25 times of physical trade(Table 4). Low and falling liquidity is also reported for the Dutch market26.US markets are recovering from a post-Enron drop in liquidity.

The insufficiency of liquidity in financial markets for hedging electricityprice risks means that investors seeking to hedge short-term price risks(or banks helping to finance these risks) cannot yet rely on these marketsto help mitigate such risks. Theoretically, investors could construct aforward price curve to help them assess these investments and to hedgetheir income stream. But financial hedging instruments appear to have amuch shorter time horizon than the term of the investment, leavingsubstantial residual risk. Indeed, it is unlikely that there would be sufficient

45


26. Newbery et al., 2003.

Electricity Ratio of Traded Amount/Market Physical Demand

Powernext (France) 0.005

EEX (Germany) 2.5

Nord Pool 8.5

England and Wales 9

Liquidity benchmark 25

Source: Montfort, 2003.

Table 4

Comparison of European Electricity Market Liquidity (as of April 2003)

46


Figure 4

Growth of Electricity Trade in Nordic Electricity Markets (1998-2003) Elspot Market

Source: Nord Pool.

0

2 000

4 000

6 000

8 000

10 000

12 000

14 000

Jan

Mar

May Ju

lSe

pN

ov Jan

Mar

May Ju

lSe

pN

ov Jan

Mar

May Ju

lSe

pN

ov Jan

Mar

May Ju

lSe

pN

ov Jan

Mar

May Ju

lSe

pN

ov Jan

Mar

May Ju

l

1998 1999 2000 2001

GWh/Month

2002 2003

demand for a long-term electricity futures contract to ever become astandard financial product.

Contract Hedges

If financial markets are not or not yet a practical tool for hedging againstshort-term risk and volatile prices, the common interest of consumersand producers in their desire to hedge against uncertain future electricityprices remains. Contract hedges between generators and consumersmight therefore provide a more promising basis for hedging the risksassociated with developing new power plants.

In fact, there appears to be relatively limited interest on the part of end-consumers in most countries to sign up for long-term contracts. Largerconsumers are likely to be more interested in long-term arrangements inorder to stabilise costs of inputs. In many markets, the existence of surplus

capacity and relatively low prices since liberalisation have discouragedlong-term contracts because of differences between producers andconsumers on future price levels. Yet, even as markets have tightened,particularly in Nord Pool, larger consumers continue to rely mainly on oneto three-year contracts rather than contracts of longer term27.

This leaves two other possibilities for longer-term contracting eitherorganisations of consumers (i.e. some form of consumer co-operative) ora contract with an electricity retailer with a large and relatively stablebase of consumers. And, in fact, there are good examples of long-termcontracts to finance new power generation investment.

The planned development of a new nuclear plant in Finland by the firmTVO is an example of a large consumer co-operative financing a new

47


27. Littlechild, 2002.

Figure 5

Growth of Electricity Trade in Nordic Electricity Markets (1998-2003) Financial Markets

0

50 000

100 000

150 000

200 000

250 000

300 000

350 000

400 000

450 000

500 000

Jan

Mar

May Ju

lSe

pN

ov Jan

Mar

May Ju

lSe

pN

ov Jan

Mar

May Ju

lSe

pN

ov Jan

Mar

May Ju

lSe

pN

ov Jan

Mar

May Ju

lSe

pN

ov Jan

Mar

May Ju

l

GWh/Month

NP Trades in GWh OTC-Clearing in GWh

1998 1999 2000 2001 2002 2003

Trans.

Note: OTC refers to over-the-counter trades. NP refers to trades made through Nord Pool.Source: Nord Pool.

power plant.TVO is a long-existing co-operative consisting of both largeelectricity consumers and some municipal utility companies.TVO, whichowns two nuclear power plants already in operation, sells electricity atcost to its investors in proportion to their contribution to theinvestment. With a guaranteed customer base who agrees to cover allcosts, TVO expects to be able to finance future investment in a powerplant at a very reasonable rate, making a nuclear plant its most cost-effective option28. In fact, it is a purchasing co-operative organisingupstream integration as a joint venture.

The development of a new co-generation power plant in the Netherlandsis a second example of a power plant financed by a 15-year power purchaseagreement between the project developer (Intergen) and the retailer(Nuon)29.The prices in the first five years of the agreement are fixed by thecontract (in turn backed by a gas supply agreement), with prices andquantities in years 6 to 15 shared between market and fixed priceagreement30.The UK retailer Centrica has in the past year entered into asimilar long-term agreement with Intergen for a UK gas-fired power plantin 200231 and more recently a four-year contract with British Energy32.

However, such long-term arrangements by Centrica and other retailersare to date unusual. It can be expected that retailers will be lessinterested in signing long-term power purchase agreements given theability of consumers to switch suppliers. This has raised concern thatlong-term contracting by retailers might be the missing link between theinvestors desire for a long-term contract and some consumers tendencyto rely on short-term markets and spot prices. The US Federal EnergyRegulatory Commission (FERC) had proposed that retailers in wholesaleelectricity markets in the United States be able to make arrangements insupply for up to three years in advance of real time. By including aforward contracting requirement in its proposals, the FERC hoped to

48

28. Santaholma, 2003.

29. Somerset, 2002.

30. The developer notes that finance was nonetheless difficult to obtain because of a strong lender aversion to themerchant component of the arrangement, and because of worsening financing conditions in general.

31. One major difference is that Centrica will supply the gas to the project. See Littlechild, 2002, p. 12.

32. British Energy, 2003.


encourage greater reliance on forward contracts rather than on themore volatile spot markets.

Another option for the consumer to hedge against shifts in power pricesis to invest in on-site distributed generation. Falling capital costs forsmaller power plants (particularly reciprocating engines or smallturbines), opportunities for economic combined heat and power (CHP)generation, and a need for higher reliability increase the feasibility andattractiveness of this option. While successful in jurisdictions with high

[Published by OECD Publishing] Power Generation (BookFi.org)

Documents

Transcript of [Published by OECD Publishing] Power Generation (BookFi.org)